5G means data center platforms must evolve

The foundation for seamless 5G implementation

5G is the next frontier in cellular connectivity, but the new generation of 5G promises more than just faster download speeds and lower latency. 5G is expected to offer higher bandwidth and more comprehensive coverage, which will open up a variety of new use cases for connectivity beyond mobile phones, from laptops and handheld Internet of Things (IoT) devices to automotive and large-scale industrial applications. According to industry forecasts, by the mid-2020s, 5G users will increase by more than 1 billion as 4G devices that are not compatible with 5G networks are migrated.

However, the transition to 5G will require significant investments in new cellular infrastructure. The 5G architecture will be a significant change from previous implementations. Driving these changes are some key features that are evolving from the fourth generation of 5G. We will review these key features and then see how these will affect the data platform architecture of a typical 5G system and explore the implementation options for different levels of the data platform and then identify the typical implementation options. In addition, an example of a mid-range data platform will be examined in detail to identify the key design choices and trade-offs.

5G feature-driven implementation architecture

Low-band 5G base stations operate in a similar frequency range as 4G mobile phones (from 600MHz to 850MHz, and provide similar range and download speeds (30Mbit/s to 250Mbit/s). As a result, low-band 5G has been phased out in many parts of the world. 5G mid-band towers use microwaves between approximately 2.5GHz and 3.7GHz, significantly increasing download speeds to the 100Mbit/s-900Mbit/s range while extending coverage by several miles. Mid-band 5G towers are already the standard in large cities and other densely populated areas, and may soon become the global standard.

High-band 5G currently operates in the 25GHz-39GHz range and offers download speeds similar to wired internet service, around 1Gbps. However, high-band 5G also has limitations. The 25GHz-39GHz range is the low end of the millimeter wave (mmW) bands. The coverage of mmWave is more limited than microwaves, which means high-band 5G will require more and smaller base stations to cover the same area as mid-band 5G.

In addition, physical obstacles such as walls or household appliances may also limit 5G high-band connections. Millimeter waves also do not handle solid objects very well. High-band 5G technology is also much more expensive than low-band technology. Therefore, in the near future, high-band 5G may be limited to large, relatively open facilities such as concert venues and stadiums.

5G Data Platform Pyramid

When determining the specific 5G layer level to be implemented, factors such as coverage, required download speeds, and cost efficiency must be considered. The 5G distributed data platform places data processing, storage, and communication at different architectural levels to optimize cost, power consumption, network performance, operating distance, and user characteristics. Closest to the edge of the network are small micro platforms that cover very small distances (tens of meters) within a building or facility. Common examples include building automation, security, and factory floor module monitoring and control. The layer above the edge device is the aggregation platform, which stitches all edge devices together and consolidates and optimizes data traffic at distances of about 100M. These devices, typically located at the building or small campus level, can analyze, filter, combine, and prioritize communications, sometimes combined with artificial intelligence (AI).

Intermediate platforms sit below the massive central datacenter (core) to provide faster responses. These responses are typically algorithm-based, selected by the core and updated regularly. These platforms provide the real-time control needed for platforms closer to the edge. Data analysis and tracking from these platforms can provide value and new revenue streams for platform providers. Cost savings from operations such as predictive maintenance, material tracking and routing, system management, and data traffic load balancing can be passed back to users (perhaps charging a subscription fee or a percentage of savings).

Big data operations run on core data center platforms. These massive data processing and storage facilities store years of historical operations and complex machine learning algorithms that provide optimization filters and processes, programming intermediate platforms for rapid response and increasing value for customers.

Evaluating different types of data center platforms

At each layer of the 5G hierarchy, designers encounter different requirements and trade-offs:

• Small platforms are the most cost, footprint, and power constrained parts of the system. They need to be easy to install and have a moderate lifecycle since multiple are needed for each aggregated platform. MCU-based implementations can be used here due to their low cost, low power consumption, and small footprint while making some minor sacrifices in flexibility and processing power.
• Aggregation platforms require great flexibility, processing power, intermediate storage, and security. Field-programmable gate array-based implementations can provide the greatest degree of flexibility and processing power, as the underlying hardware can be reprogrammed as needed for new protocols, new AI algorithms, or new value additions to customers.
  FPGAs also scale easily, allowing vendors to create different product tiers with varying levels of flexibility and processing power at different price and value points.
• The middle platform requires the highest levels of raw processing power, security, and flexibility. Cost, power consumption, and footprint are acceptable sacrifices. At this level, a hybrid combination of memory protection units (for raw processing power for mundane operations) and FPGAs (for flexibility and adaptability) would be the best approach. FPGAs can even be reprogrammed in real time as needed to respond to growing demand for important functions such as video processing, encryption, decryption, searching, and filtering. AI algorithms can even anticipate this demand by analyzing clues found in key indicators such as occupancy, traffic patterns, weather, and more.

How does the aggregation platform work?

Let’s look at a mid-range feature aggregation platform and see how key functions are implemented in a device. Utilizing an FPGA and an on-chip microcontroller allows the device to operate from the MCU, providing a known secure starting point for booting and secure updates. This root of trust uses reliable and protected encryption, decryption, and secure key storage operations to prevent hackers and viruses from taking over the system. The MCU can also handle standard operations such as communications, packet processing, video processing, compression, and storage efficiency. On-chip FPGA hardware can be used for less common but high-power operations to keep the MCU free for common operations that must be completed in a timely manner.

FPGAs can be used for digital filtering, image processing, image recognition, and similar specialized operations, perhaps combined with AI and machine learning algorithms to predict and program the hardware on an as-needed basis. Over time, new algorithms can be identified, created, and downloaded from the intermediate and core platforms to further optimize performance, creating new revenue streams for platform providers and cost savings for customers.

5G will mean the convergence of cloud, core, and edge. However, it will be important that each component runs on, or has access to, the right type of infrastructure. The core will always be needed, even with the importance of edge computing for 5G, but as 5G connections proliferate and the number of 5G-enabled devices continues to grow, the need for small to medium-sized distributed data center platforms for these devices will increase. Computing from core to edge will be the hallmark of a 5G-enabled world. It will be a world where thermostats, refrigerators, planes, trains, and cars are all connected to the same cellular network as your phone.